Method and apparatus for performing machine translation using a unified language model and translation model

Abstract

The present invention is a method and apparatus for processing a phrase in a first language for translation to a second language. A plurality of possible linguistic patterns are identified in the second language, that correspond to the phrase in the first language. For each of the patterns identified, a probability for the pattern is calculated, based on a combination of the language model probability for the pattern and a translation model probability for the pattern. In one embodiment, an output is also provided which is indicative of a translation of the phrase in the first language to the second language based upon the translation probabilities calculated for the patterns.

1. A computer-implemented method of processing a phrase in a first language for translation to a second language, comprising:
 receiving the phrase in the first language; 
 identifying a plurality of possible linguistic patterns in the second language associated with the phrase in the first language, wherein each of the plurality of possible linguistic patterns represents a grouping of components relative to the phrase; and 
 for each pattern, calculating a translation probability for the pattern based on a combination of a language model probability for the pattern and a translation model probability for the pattern. 
 
   
   
     2. The method of  claim 1  and further comprising:
 identifying a highest translation probability calculated; and 
 identifying a linguistic pattern, for which the highest translation probability was calculated, as indicative of a likely phrase translation of the phrase in the first language. 
 
   
   
     3. The method of  claim 2  and further comprising:
 providing an output as a translation of the phrase in the first language to the second language based on the linguistic pattern identified. 
 
   
   
     4. The method of  claim 1  wherein identifying a plurality of possible linguistic patterns, comprises:
 accessing a bilingual data store that includes linguistic patterns in the second language associated with phrases in the first language. 
 
   
   
     5. The method of  claim 1  wherein calculating a translation probability further comprises:
 calculating a pattern probability for the pattern. 
 
   
   
     6. A computer-implemented method of processing a multi-word phrase in a first language for translation to a second language, comprising:
 receiving the multi-word phrase in the first language; 
 identifying a plurality of possible linguistic patterns in the second language that correspond to the phrase in the first language, wherein each of the plurality of possible linguistic patterns represents a grouping of translation components relative to the phrase; and 
 calculating a translation probability for translation of the multi-word phrase in the first language to one of the plurality of linguistic patterns in the second language. 
 
   
   
     7. The method of  claim 6  wherein calculating a translation probability comprises:
 for each of the linguistic patterns identified, calculating the translation probability as a combination of a language model probability for the pattern in the second language and as a translation model probability for the phrase in the first language, given the linguistic pattern in the second language. 
 
   
   
     8. The method of  claim 7  wherein calculating a translation probability further comprises:
 calculating the translation probability based on a pattern probability for the linguistic pattern. 
 
   
   
     9. The method of  claim 7  and further comprising:
 identifying a highest translation probability calculated; and 
 identifying a linguistic pattern, for which the highest translation probability was calculated, as indicative of a likely phrase translation of the phrase in the first language. 
 
   
   
     10. The method of  claim 9  and further comprising:
 providing an output as a translation of the phrase in the first language to the second language based on the linguistic pattern identified. 
 
   
   
     11. The method of  claim 7  wherein identifying a plurality of possible linguistic patterns, comprises:
 accessing a bilingual data store that includes linguistic patterns in the second language associated with phrases in the first language. 
 
   
   
     12. A natural language processing system, comprising:
 a pattern engine receiving a phrase in a first language and identifying a plurality of linguistic patterns in a second language, associated with the phrase in the first language, possibly corresponding to a translation of the phrase from the first language to the second language, wherein each of the plurality of linguistic patterns represents a grouping of components relative to the phrase; and 
 a probability generator configured to generate, for each linguistic pattern identified, a translation probability for translating the phrase in the first language to the second language in the linguistic pattern. 
 
   
   
     13. The system of  claim 12  wherein the pattern engine, comprises:
 a bi-lingual data store storing phrases in the first language and corresponding linguistic patterns in the second language. 
 
   
   
     14. The system of  claim 13  wherein the probability generator comprises:
 a translation model, the probability generator being configured to generate the translation probability by accessing the translation model. 
 
   
   
     15. The system of  claim 14  wherein the probability generator further comprises:
 a language model in the second language, the probability generator being configured to generate the translation probability by accessing the language model. 
 
   
   
     16. The system of  claim 15  wherein the probability generator is configured to:
 identify a highest translation probability calculated; and 
 identify a linguistic pattern, for which the highest translation probability was calculated, as indicative of a likely phrase translation of the phrase in the first language.